The major purpose of the present study is to investigate the relative efficacy of three alternative treatment models for homeless poly-addicted men. Using a true experimental design, 700 subjects will be randomly assigned over a 16 month period to three treatment conditions: typical shelter services with minimal casework (Group 1); shelter with intensive case management using a community network of services (Group 2); and intensive case management with integrated, comprehensive services provided at one site (Group 3). Outcome assessment will focus on the effectiveness of these approaches in reducing drug and alcohol use, fostering placement in independent living, promoting residential stability, increasing economic and employment status, improving family and social relations, and enhancing mental health. In addition, the study will investigate which service utilization patterns and which personal factors (demographic, social, and psychological) best predict success and failure of treatment. Data will be collected at intake, discharge, and 6 months after discharge. Distinctive features of this project include the focus on minority, male, homeless, polydrug users; the use of indigenous case managers; the provision of aftercare services following discharge; a comparison between two innovative intensive case management models; and a combination of multiple assessment methods.